{ "epoch": 1.0, "eval_logits/chosen": -3.9838693141937256, "eval_logits/rejected": -4.009171962738037, "eval_logps/chosen": -544.685791015625, "eval_logps/rejected": -438.45953369140625, "eval_loss": 0.5608103275299072, "eval_rewards/accuracies": 0.6930000185966492, "eval_rewards/chosen": 0.4690808057785034, "eval_rewards/margins": 0.5145381689071655, "eval_rewards/rejected": -0.04545731097459793, "eval_runtime": 146.3889, "eval_samples": 2000, "eval_samples_per_second": 13.662, "eval_steps_per_second": 1.708 }